One "lump" or two? Are there independent or common orthographic lexicons for reading and spelling?
Bartha, Michael Christopher
Doctor of Philosophy
Three theories on the relationship between the lexical orthographic representations utilized in reading and spelling can be distinguished. They are a common or single lexicon theory, an independent lexicons theory, and a "separate but linked" lexicons theory. Little research with normal populations has addressed these issues. Neuropsychological studies have produced mixed results. Three experiments examined the validity of each of these theories utilizing both long-term and short-term priming paradigms. Experiment 1 examined the relationship between reading and spelling utilizing a perceptual identification task, and found facilitation in terms of accuracy of later perceptual identification performance only for previously read items. Experiment 2a found marginally significant facilitation of later spelling performance from previously spelling an item at study in a spelling-probe task but no significant facilitation in a missing-letter task. Experiment 2b found significant facilitation from previously spelling a word at study to later performance on a spelling-probe task. However, this facilitation was only significant across subjects and not items. Experiment 3 examined the influence of spelling on reading and reading on spelling utilizing a short-term priming paradigm. Subjects performed either a spelling-probe or a lexical decision task, and facilitation was measured to repeated items that appeared with one, two, five, and ten items intervening between the first and the second presentations. Significant facilitation was observed from reading to reading, and spelling to spelling, as well as from spelling to reading. No significant priming was obtained from reading to spelling. Read primes produced significantly more priming than spelled primes for read targets. There was no corresponding, statistically significant, advantage for spelled primes over read primes for spelled targets. No significant interactions with the number of intervening items (lag) on the amount of priming were observed. These data are interpreted as supportive of a lexical orthographic system which involves two distinct lexicons, (one for input and one for output) which are linked together in some manner.
Linguistics; Experimental psychology; Cognitive psychology